DOCSLIB.ORG

Partitions of Unity and Paracompactness

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Partitions of Unity and Paracompactness Math 535 - General Topology Additional notes Martin Frankland November 12, 2012 1 Partitions of unity Definition 1.1. Let X be a topological space and f : X ! R a continuous function. The support of f is the closed subset supp f := fx 2 X j f(x) 6= 0g ⊆ X: Definition 1.2. A cover fUαgα2A of a space X is locally finite if for all x 2 X, there is a neighborhood Nx of x that intersects only finitely many of the Uα, i.e. Nx \ Uα 6= ; for finitely many α 2 A. Note that the definition applies to any kind of cover, not just open covers. Definition 1.3. A partition of unity on a space X is a family of continuous functions fρβ : X ! [0; 1]gβ2B satisfying the following two properties. 1. The family fsupp ρβgβ2B is locally finite. P 2. β2B ρβ(x) = 1 for all x 2 X. In particular, the supports fsupp ρβgβ2B form a cover of X, and for all x 2 X, ρβ(x) = 0 for all except finitely many indices β 2 B. The partition of unity is subordinate to a cover fUαgα2A of X if for all β 2 B, there is an index α = α(β) 2 A satisfying supp ρβ ⊆ Uα(β): Remark 1.4. One could assume WLOG that the indexing set B is A. Indeed, one could add together the functions ρβ supported on the same Uα, and consider the constant zero function for each index α that does not appear as α(β). It is useful to have partitions of unity subordinate to any open cover. We will find conditions on a space that make this always possible. 1 2 Paracompactness Definition 2.1. Let U be a cover of X.A refinement of U is a cover V of X such that every member V 2 V is a subset of some U 2 U. In other words, a cover fVβgβ2B is a refinement of the cover fUαgα2A if for all β 2 B, there is an index α(β) 2 A satisfying Vβ ⊆ Uα(β): Example 2.2. f(n; n + 2)gn2Z is a refinement of the cover fRg of R. 1 Example 2.3. f(0; 1)g is a refinement of the cover f(n; 1)gn=0 of (0; 1). Definition 2.4. A topological space X is paracompact if every open cover of X admits a locally finite open refinement. Remark 2.5. Some authors (e.g. Bredon, Willard) include the Hausdorff condition in the defi- nition of paracompact. Other authors (e.g. Munkres, Wikipedia) do not assume that the space is Hausdorff. We will not assume that the definition of paracompact includes Hausdorff. Example 2.6. Every compact space is paracompact. Example 2.7. Rn is paracompact. This will follow from 2.13. Definition 2.8. A space is σ-compact if it is a countable union of compact subspaces. Example 2.9. n is σ-compact, as it is the union of closed balls n = S B (0). R R k2N k Example 2.10. Any closed subset of Rn is σ-compact. Example 2.11. More generally, any closed subset C ⊆ X of a σ-compact Hausdorff space X is σ-compact. Example 2.12. Any second-countable manifold is σ-compact. More generally, any second- countable locally compact space is σ-compact (c.f. Homework 12 #2). Proposition 2.13. Every locally compact, σ-compact, Hausdorff space is paracompact. Proposition 2.14. A closed subspace of a paracompact space is paracompact. Proof. Homework 12 #3. Lemma 2.15. If fAigi2I is a locally finite collection of subsets Ai ⊆ X, then we have [ [ Ai = Ai: i2I i2I Proposition 2.16. Every paracompact Hausdorff space is normal. Proposition 2.17 (Shrinking lemma). Let X be a paracompact Hausdorff space and fUαgα2A an open cover. Then there is a locally finite open cover fVαgα2A satisfying Vα ⊆ Uα for all α 2 A. Note that the indexing set A is the same for both open covers. Note also that Vα is allowed to be empty. Theorem 2.18 (Existence of partitions of unity). Let X be a paracompact Hausdorff space and U = fUαgα2A an open cover. Then there is a partition of unity on X subordinate to U. 2 3 Applications to manifolds Theorem 3.1. Any compact Hausdorff manifold can be embedded in RN for some N. Proof. Let M be an m-dimensional compact Hausdorff manifold. For each x 2 M, a coordinate m chart about x consists of a homeomorphism 'x : Ux ! Vx ⊆ R where Ux is an open neigh- m borhood of x and Vx is an open subset of R . Since M is compact, the open cover fUxgx2M admits a finite subcover fU1;:::;Ukg. Since M is paracompact and Hausdorff, there exists a k partition of unity fρigi=1 with supp ρi ⊆ Ui. m For i = 1; : : : ; k, define maps hi : M ! R by ( ρi(x)'i(x) if x 2 Ui hi(x) = 0 otherwise. These maps hi are well defined and continuous. Now define the continuous map k k z }| { z m }| m{ ∼ k(m+1) g : M ! R × ::: × R × R × ::: × R = R x 7! (ρ1(x); : : : ; ρk(x); h1(x); : : : ; hk(x)) : Since M is compact and RN is Hausdorff, g is a closed map. To show that g is an embedding, it remains to show that g is injective. Assume g(x) = g(y) for some x; y 2 M. Then ρi(x) = ρi(y) for all i. Let j be an index where ρj(x) > 0 and thus ρj(y) = ρj(x) > 0. We obtain hj(x) = hj(y) ρj(x)'j(x) = ρj(y)'j(y) ) 'j(x) = 'j(y) ) x = y since 'j is injective. This example illustrates how partitions of unity on a manifold can be useful. Since many interesting manifolds are not compact, it would be useful to know when a (Hausdorff) manifold is paracompact. Theorem 3.2. Let M be a Hausdorff manifold. Then M is paracompact if and only if each connected component of M is second-countable. Proof. Let fMigi2I be the connected components of M. Recall that manifolds are locally path- ` connected. Therefore M = i2I Mi is the coproduct of its connected components (which are the same as its path components). (() Note that any manifold is locally compact. Since Mi is locally compact and second- countable, it is σ-compact (by 2.12). Since Mi is locally compact, σ-compact, and Hausdorff, it is paracompact (by 2.13). Since ` M = i2I Mi is a coproduct of paracompact spaces, it is paracompact (c.f. Homework 12 #4). 3 ()) Since Mi is connected, locally compact, paracompact, and Hausdorff, it is σ-compact (by 3.3). Since Mi is σ-compact and every point x 2 Mi has a second-countable neighborhood, Mi is second-countable (by 3.4). Proposition 3.3. Let X be a connected, locally compact, paracompact, Hausdorff space. Then X is σ-compact. Proposition 3.4. Let X be a σ-compact space such that every point x 2 X has a second- countable neighborhood. Then X is second-countable. Here is another application of partitions of unity. Proposition 3.5. Every paracompact smooth manifold admits a Riemannian metric. Yet another application is in defining integration on manifolds. One can define integration within a coordinate chart, and then on the entire manifold using a partition of unity. 4.
Load more

Recommended publications
  • Distributions:The Evolutionof a Mathematicaltheory
    Historia Mathematics 10 (1983) 149-183 DISTRIBUTIONS:THE EVOLUTIONOF A MATHEMATICALTHEORY BY JOHN SYNOWIEC INDIANA UNIVERSITY NORTHWEST, GARY, INDIANA 46408 SUMMARIES The theory of distributions, or generalized func- tions, evolved from various concepts of generalized solutions of partial differential equations and gener- alized differentiation. Some of the principal steps in this evolution are described in this paper. La thgorie des distributions, ou des fonctions g&&alis~es, s'est d&eloppeg 2 partir de divers con- cepts de solutions g&&alis6es d'gquations aux d&i- &es partielles et de diffgrentiation g&&alis6e. Quelques-unes des principales &apes de cette &olution sont d&rites dans cet article. Die Theorie der Distributionen oder verallgemein- erten Funktionen entwickelte sich aus verschiedenen Begriffen von verallgemeinerten Lasungen der partiellen Differentialgleichungen und der verallgemeinerten Ableitungen. In diesem Artikel werden einige der wesentlichen Schritte dieser Entwicklung beschrieben. 1. INTRODUCTION The founder of the theory of distributions is rightly con- sidered to be Laurent Schwartz. Not only did he write the first systematic paper on the subject [Schwartz 19451, which, along with another paper [1947-19481, already contained most of the basic ideas, but he also wrote expository papers on distributions for electrical engineers [19481. Furthermore, he lectured effec- tively on his work, for example, at the Canadian Mathematical Congress [1949]. He showed how to apply distributions to partial differential equations [19SObl and wrote the first general trea- tise on the subject [19SOa, 19511. Recognition of his contri- butions came in 1950 when he was awarded the Fields' Medal at the International Congress of Mathematicians, and in his accep- tance address to the congress Schwartz took the opportunity to announce his celebrated "kernel theorem" [195Oc].
    [Show full text]
  • [Math.DG] 1 Feb 2002
    The BIC of a conical fibration. Martintxo Saralegi-Aranguren∗ Robert Wolak† Universit´ed’Artois Uniwersytet Jagiellonski November 14, 2018 Abstract In the paper we introduce the notions of a singular fibration and a singular Seifert fi- bration. These notions are natural generalizations of the notion of a locally trivial fibration to the category of stratified pseudomanifolds. For singular foliations defined by such fibra- tions we prove a de Rham type theorem for the basic intersection cohomology introduced the authors in a recent paper. One of important examples of such a structure is the natural projection onto the leaf space for the singular Riemannian foliation defined by an action of a compact Lie group on a compact smooth manifold. The failure of the Poincar´eduality for the homology and cohomology of some singular spaces led Goresky and MacPherson to introduce a new homology theory called intersection homology which took into account the properties of the singularities of the considered space (cf. [10]). This homology is defined for stratified pseudomanifolds. The initial idea was generalized in several ways. The theories of simplicial and singular homologies were developed as well as weaker notions of the perversity were proposed (cf. [11, 12]). Several versions of the Poincar´eduality were proved taking into account the notion of dual perversities. Finally, the deRham intersection cohomology was defined by Goresky and MacPherson for Thom-Mather stratified spaces. The first written reference is the paper by J.L. Brylinski, cf. [6]. This led to the search for the ”de Rham - type” theorem for the intersection cohomology. The first author in his thesis and several subsequent publications, (cf.
    [Show full text]
  • K-THEORY of FUNCTION RINGS Theorem 7.3. If X Is A
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Journal of Pure and Applied Algebra 69 (1990) 33-50 33 North-Holland K-THEORY OF FUNCTION RINGS Thomas FISCHER Johannes Gutenberg-Universitit Mainz, Saarstrage 21, D-6500 Mainz, FRG Communicated by C.A. Weibel Received 15 December 1987 Revised 9 November 1989 The ring R of continuous functions on a compact topological space X with values in IR or 0Z is considered. It is shown that the algebraic K-theory of such rings with coefficients in iZ/kH, k any positive integer, agrees with the topological K-theory of the underlying space X with the same coefficient rings. The proof is based on the result that the map from R6 (R with discrete topology) to R (R with compact-open topology) induces a natural isomorphism between the homologies with coefficients in Z/kh of the classifying spaces of the respective infinite general linear groups. Some remarks on the situation with X not compact are added. 0. Introduction For a topological space X, let C(X, C) be the ring of continuous complex valued functions on X, C(X, IR) the ring of continuous real valued functions on X. KU*(X) is the topological complex K-theory of X, KO*(X) the topological real K-theory of X. The main goal of this paper is the following result: Theorem 7.3. If X is a compact space, k any positive integer, then there are natural isomorphisms: K,(C(X, C), Z/kZ) = KU-‘(X, Z’/kZ), K;(C(X, lR), Z/kZ) = KO -‘(X, UkZ) for all i 2 0.
    [Show full text]
  • Arxiv:0704.2891V3 [Math.AG] 5 Dec 2007 Schwartz Functions on Nash Manifolds
    Schwartz functions on Nash manifolds Avraham Aizenbud and Dmitry Gourevitch ∗ July 11, 2011 Abstract In this paper we extend the notions of Schwartz functions, tempered func- tions and generalized Schwartz functions to Nash (i.e. smooth semi-algebraic) manifolds. We reprove for this case classically known properties of Schwartz functions on Rn and build some additional tools which are important in rep- resentation theory. Contents 1 Introduction 2 1.1 Mainresults................................ 3 1.2 Schwartz sections of Nash bundles . 4 1.3 Restricted topologyand sheaf properties . .... 4 1.4 Possibleapplications ........................... 5 1.5 Summary ................................. 6 1.6 Remarks.................................. 6 2 Semi-algebraic geometry 8 2.1 Basicnotions ............................... 8 arXiv:0704.2891v3 [math.AG] 5 Dec 2007 2.2 Tarski-Seidenberg principle of quantifier elimination anditsapplications............................ 8 2.3 Additional preliminary results . .. 10 ∗Avraham Aizenbud and Dmitry Gourevitch, Faculty of Mathematics and Computer Science, The Weizmann Institute of Science POB 26, Rehovot 76100, ISRAEL. E-mails: [email protected], [email protected]. Keywords: Schwartz functions, tempered functions, generalized functions, distributions, Nash man- ifolds. 1 3 Nash manifolds 11 3.1 Nash submanifolds of Rn ......................... 11 3.2 Restricted topological spaces and sheaf theory over them........ 12 3.3 AbstractNashmanifolds . 14 3.3.1 ExamplesandRemarks. 14 3.4 Nashvectorbundles ........................... 15 3.5 Nashdifferentialoperators . 16 3.5.1 Algebraic differential operators on a Nash manifold . .... 17 3.6 Nashtubularneighborhood . 18 4 Schwartz and tempered functions on affine Nash manifolds 19 4.1 Schwartzfunctions ............................ 19 4.2 Temperedfunctions. .. .. .. 20 4.3 Extension by zero of Schwartz functions . .. 20 4.4 Partitionofunity ............................. 21 4.5 Restriction and sheaf property of tempered functions .
    [Show full text]
  • Arxiv:1404.7630V2 [Math.AT] 5 Nov 2014 Asoffsae,Se[S4 P8,Vr6.Isedof Instead Ver66]
    PROPER BASE CHANGE FOR SEPARATED LOCALLY PROPER MAPS OLAF M. SCHNÜRER AND WOLFGANG SOERGEL Abstract. We introduce and study the notion of a locally proper map between topological spaces. We show that fundamental con- structions of sheaf theory, more precisely proper base change, pro- jection formula, and Verdier duality, can be extended from contin- uous maps between locally compact Hausdorff spaces to separated locally proper maps between arbitrary topological spaces. Contents 1. Introduction 1 2. Locally proper maps 3 3. Proper direct image 5 4. Proper base change 7 5. Derived proper direct image 9 6. Projection formula 11 7. Verdier duality 12 8. The case of unbounded derived categories 15 9. Remindersfromtopology 17 10. Reminders from sheaf theory 19 11. Representability and adjoints 21 12. Remarks on derived functors 22 References 23 arXiv:1404.7630v2 [math.AT] 5 Nov 2014 1. Introduction The proper direct image functor f! and its right derived functor Rf! are defined for any continuous map f : Y → X of locally compact Hausdorff spaces, see [KS94, Spa88, Ver66]. Instead of f! and Rf! we will use the notation f(!) and f! for these functors. They are embedded into a whole collection of formulas known as the six-functor-formalism of Grothendieck. Under the assumption that the proper direct image functor f(!) has finite cohomological dimension, Verdier proved that its ! derived functor f! admits a right adjoint f . Olaf Schnürer was supported by the SPP 1388 and the SFB/TR 45 of the DFG. Wolfgang Soergel was supported by the SPP 1388 of the DFG. 1 2 OLAFM.SCHNÜRERANDWOLFGANGSOERGEL In this article we introduce in 2.3 the notion of a locally proper map between topological spaces and show that the above results even hold for arbitrary topological spaces if all maps whose proper direct image functors are involved are locally proper and separated.
    [Show full text]
  • 1.1.5 Hausdorff Spaces
    1. Preliminaries Proof. Let xn n N be a sequence of points in X convergent to a point x X { } 2 2 and let U (f(x)) in Y . It is clear from Definition 1.1.32 and Definition 1.1.5 1 2F that f − (U) (x). Since xn n N converges to x,thereexistsN N s.t. 1 2F { } 2 2 xn f − (U) for all n N.Thenf(xn) U for all n N. Hence, f(xn) n N 2 ≥ 2 ≥ { } 2 converges to f(x). 1.1.5 Hausdor↵spaces Definition 1.1.40. A topological space X is said to be Hausdor↵ (or sepa- rated) if any two distinct points of X have neighbourhoods without common points; or equivalently if: (T2) two distinct points always lie in disjoint open sets. In literature, the Hausdor↵space are often called T2-spaces and the axiom (T2) is said to be the separation axiom. Proposition 1.1.41. In a Hausdor↵space the intersection of all closed neigh- bourhoods of a point contains the point alone. Hence, the singletons are closed. Proof. Let us fix a point x X,whereX is a Hausdor↵space. Denote 2 by C the intersection of all closed neighbourhoods of x. Suppose that there exists y C with y = x. By definition of Hausdor↵space, there exist a 2 6 neighbourhood U(x) of x and a neighbourhood V (y) of y s.t. U(x) V (y)= . \ ; Therefore, y/U(x) because otherwise any neighbourhood of y (in particular 2 V (y)) should have non-empty intersection with U(x).
    [Show full text]
  • MATH 418 Assignment #7 1. Let R, S, T Be Positive Real Numbers Such That R
    MATH 418 Assignment #7 1. Let r, s, t be positive real numbers such that r + s + t = 1. Suppose that f, g, h are nonnegative measurable functions on a measure space (X, S, µ). Prove that Z r s t r s t f g h dµ ≤ kfk1kgk1khk1. X s t Proof . Let u := g s+t h s+t . Then f rgsht = f rus+t. By H¨older’s inequality we have Z Z rZ s+t r s+t r 1/r s+t 1/(s+t) r s+t f u dµ ≤ (f ) dµ (u ) dµ = kfk1kuk1 . X X X For the same reason, we have Z Z s t s t s+t s+t s+t s+t kuk1 = u dµ = g h dµ ≤ kgk1 khk1 . X X Consequently, Z r s t r s+t r s t f g h dµ ≤ kfk1kuk1 ≤ kfk1kgk1khk1. X 2. Let Cc(IR) be the linear space of all continuous functions on IR with compact support. (a) For 1 ≤ p < ∞, prove that Cc(IR) is dense in Lp(IR). Proof . Let X be the closure of Cc(IR) in Lp(IR). Then X is a linear space. We wish to show X = Lp(IR). First, if I = (a, b) with −∞ < a < b < ∞, then χI ∈ X. Indeed, for n ∈ IN, let gn the function on IR given by gn(x) := 1 for x ∈ [a + 1/n, b − 1/n], gn(x) := 0 for x ∈ IR \ (a, b), gn(x) := n(x − a) for a ≤ x ≤ a + 1/n and gn(x) := n(b − x) for b − 1/n ≤ x ≤ b.
    [Show full text]
  • The Non-Hausdorff Number of a Topological Space
    THE NON-HAUSDORFF NUMBER OF A TOPOLOGICAL SPACE IVAN S. GOTCHEV Abstract. We call a non-empty subset A of a topological space X finitely non-Hausdorff if for every non-empty finite subset F of A and every family fUx : x 2 F g of open neighborhoods Ux of x 2 F , \fUx : x 2 F g 6= ; and we define the non-Hausdorff number nh(X) of X as follows: nh(X) := 1 + supfjAj : A is a finitely non-Hausdorff subset of Xg. Using this new cardinal function we show that the following three inequalities are true for every topological space X d(X) (i) jXj ≤ 22 · nh(X); 2d(X) (ii) w(X) ≤ 2(2 ·nh(X)); (iii) jXj ≤ d(X)χ(X) · nh(X) and (iv) jXj ≤ nh(X)χ(X)L(X) is true for every T1-topological space X, where d(X) is the density, w(X) is the weight, χ(X) is the character and L(X) is the Lindel¨ofdegree of the space X. The first three inequalities extend to the class of all topological spaces d(X) Pospiˇsil'sinequalities that for every Hausdorff space X, jXj ≤ 22 , 2d(X) w(X) ≤ 22 and jXj ≤ d(X)χ(X). The fourth inequality generalizes 0 to the class of all T1-spaces Arhangel ski˘ı’s inequality that for every Hausdorff space X, jXj ≤ 2χ(X)L(X). It is still an open question if 0 Arhangel ski˘ı’sinequality is true for all T1-spaces. It follows from (iv) that the answer of this question is in the afirmative for all T1-spaces with nh(X) not greater than the cardianilty of the continuum.
    [Show full text]
  • Professor Smith Math 295 Lecture Notes
    Professor Smith Math 295 Lecture Notes by John Holler Fall 2010 1 October 29: Compactness, Open Covers, and Subcovers. 1.1 Reexamining Wednesday's proof There seemed to be a bit of confusion about our proof of the generalized Intermediate Value Theorem which is: Theorem 1: If f : X ! Y is continuous, and S ⊆ X is a connected subset (con- nected under the subspace topology), then f(S) is connected. We'll approach this problem by first proving a special case of it: Theorem 1: If f : X ! Y is continuous and surjective and X is connected, then Y is connected. Proof of 2: Suppose not. Then Y is not connected. This by definition means 9 non-empty open sets A; B ⊂ Y such that A S B = Y and A T B = ;. Since f is continuous, both f −1(A) and f −1(B) are open. Since f is surjective, both f −1(A) and f −1(B) are non-empty because A and B are non-empty. But the surjectivity of f also means f −1(A) S f −1(B) = X, since A S B = Y . Additionally we have f −1(A) T f −1(B) = ;. For if not, then we would have that 9x 2 f −1(A) T f −1(B) which implies f(x) 2 A T B = ; which is a contradiction. But then these two non-empty open sets f −1(A) and f −1(B) disconnect X. QED Now we want to show that Theorem 2 implies Theorem 1. This will require the help from a few lemmas, whose proofs are on Homework Set 7: Lemma 1: If f : X ! Y is continuous, and S ⊆ X is any subset of X then the restriction fjS : S ! Y; x 7! f(x) is also continuous.
    [Show full text]
  • Paracompactness and C2 -Paracompactness
    Turkish Journal of Mathematics Turk J Math (2019) 43: 9 – 20 http://journals.tubitak.gov.tr/math/ © TÜBİTAK Research Article doi:10.3906/mat-1804-54 C -Paracompactness and C2 -paracompactness Maha Mohammed SAEED∗,, Lutfi KALANTAN,, Hala ALZUMI, Department of Mathematics, King Abdulaziz University, Jeddah, Saudi Arabia Received: 17.04.2018 • Accepted/Published Online: 29.08.2018 • Final Version: 18.01.2019 Abstract: A topological space X is called C -paracompact if there exist a paracompact space Y and a bijective function f : X −! Y such that the restriction fjA : A −! f(A) is a homeomorphism for each compact subspace A ⊆ X .A topological space X is called C2 -paracompact if there exist a Hausdorff paracompact space Y and a bijective function f : X −! Y such that the restriction fjA : A −! f(A) is a homeomorphism for each compact subspace A ⊆ X . We investigate these two properties and produce some examples to illustrate the relationship between them and C -normality, minimal Hausdorff, and other properties. Key words: Normal, paracompact, C -paracompact, C2 -paracompact, C -normal, epinormal, mildly normal, minimal Hausdorff, Fréchet, Urysohn 1. Introduction We introduce two new topological properties, C -paracompactness and C2 -paracompactness. They were defined by Arhangel’skiĭ. The purpose of this paper is to investigate these two properties. Throughout this paper, we denote an ordered pair by hx; yi, the set of positive integers by N, the rational numbers by Q, the irrational numbers by P, and the set of real numbers by R. T2 denotes the Hausdorff property. A T4 space is a T1 normal space and a Tychonoff space (T 1 ) is a T1 completely regular space.
    [Show full text]
  • Topological Description of Riemannian Foliations with Dense Leaves
    TOPOLOGICAL DESCRIPTION OF RIEMANNIAN FOLIATIONS WITH DENSE LEAVES J. A. AL¶ VAREZ LOPEZ*¶ AND A. CANDELy Contents Introduction 1 1. Local groups and local actions 2 2. Equicontinuous pseudogroups 5 3. Riemannian pseudogroups 9 4. Equicontinuous pseudogroups and Hilbert's 5th problem 10 5. A description of transitive, compactly generated, strongly equicontinuous and quasi-e®ective pseudogroups 14 6. Quasi-analyticity of pseudogroups 15 References 16 Introduction Riemannian foliations occupy an important place in geometry. An excellent survey is A. Haefliger’s Bourbaki seminar [6], and the book of P. Molino [13] is the standard reference for riemannian foliations. In one of the appendices to this book, E. Ghys proposes the problem of developing a theory of equicontinuous foliated spaces parallel- ing that of riemannian foliations; he uses the suggestive term \qualitative riemannian foliations" for such foliated spaces. In our previous paper [1], we discussed the structure of equicontinuous foliated spaces and, more generally, of equicontinuous pseudogroups of local homeomorphisms of topological spaces. This concept was di±cult to develop because of the nature of pseudogroups and the failure of having an in¯nitesimal characterization of local isome- tries, as one does have in the riemannian case. These di±culties give rise to two versions of equicontinuity: a weaker version seems to be more natural, but a stronger version is more useful to generalize topological properties of riemannian foliations. Another relevant property for this purpose is quasi-e®ectiveness, which is a generalization to pseudogroups of e®ectiveness for group actions. In the case of locally connected fo- liated spaces, quasi-e®ectiveness is equivalent to the quasi-analyticity introduced by Date: July 19, 2002.
    [Show full text]
  • Base-Paracompact Spaces
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Topology and its Applications 128 (2003) 145–156 www.elsevier.com/locate/topol Base-paracompact spaces John E. Porter Department of Mathematics and Statistics, Faculty Hall Room 6C, Murray State University, Murray, KY 42071-3341, USA Received 14 June 2001; received in revised form 21 March 2002 Abstract A topological space is said to be totally paracompact if every open base of it has a locally finite subcover. It turns out this property is very restrictive. In fact, the irrationals are not totally paracompact. Here we give a generalization, base-paracompact, to the notion of totally paracompactness, and study which paracompact spaces satisfy this generalization. 2002 Elsevier Science B.V. All rights reserved. Keywords: Paracompact; Base-paracompact; Totally paracompact 1. Introduction A topological space X is paracompact if every open cover of X has a locally finite open refinement. A topological space is said to be totally paracompact if every open base of it has a locally finite subcover. Corson, McNinn, Michael, and Nagata announced [3] that the irrationals have a basis which does not have a locally finite subcover, and hence are not totally paracompact. Ford [6] was the first to give the definition of totally paracompact spaces. He used this property to show the equivalence of the small and large inductive dimension in metric spaces. Ford also showed paracompact, locally compact spaces are totally paracompact. Lelek [9] studied totally paracompact and totally metacompact separable metric spaces. He showed that these two properties are equivalent in separable metric spaces.
    [Show full text]